A Textless Metric for Speech-to-Speech Comparison

Autor: Besacier, Laurent, Ribeiro, Swen, Galibert, Olivier, Calapodescu, Ioan

In this paper, we introduce a new and simple method for comparing speech utterances without relying on text transcripts. Our speech-to-speech comparison metric utilizes state-of-the-art speech2unit encoders like HuBERT to convert speech utterances in

Externí odkaz: http://arxiv.org/abs/2210.11835

Towards lifelong human assisted speaker diarization

Autor: Shamsi, Meysam, Larcher, Anthony, Barrault, Loic, Meignier, Sylvain, Prokopalo, Yevheni, Tahon, Marie, Mehrish, Ambuj, Petitrenaud, Simon, Galibert, Olivier, Gaist, Samuel, Anjos, André, Marcel, Sebastien, Costa-jussà, Marta R.

Publikováno v: In Computer Speech & Language January 2023 77

ASR Performance Prediction on Unseen Broadcast Programs using Convolutional Neural Networks

Autor: Elloumi, Zied, Besacier, Laurent, Galibert, Olivier, Kahn, Juliette, Lecouteux, Benjamin

In this paper, we address a relatively new task: prediction of ASR performance on unseen broadcast programs. We first propose an heterogenous French corpus dedicated to this task. Two prediction approaches are compared: a state-of-the-art performance

Externí odkaz: http://arxiv.org/abs/1804.08477

Le community management au service de la médiation des savoirs : une instrumentalisation de l’intelligence collective

Autor: Galibert, Olivier

Publikováno v: Les Cahiers du numérique
Les Cahiers du numérique, Lavoisier, A paraître, Intelligences collectives : communautés et interactions épistémiques

Dans cet article, nous interrogerons l’émergence et la diffusion du Community Management dans les processus de médiation des savoirs partagés et distribués en ligne. Nous émettons l’hypothèse que les pratiques « métiers » des Community M

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::5dcdad018ef85a24c9b0b991f4852958
https://doi.org/10.3166/lcn.2021.004

Analyzing BERT Cross-lingual Transfer Capabilities in Continual Sequence Labeling

Autor: Coria, Juan Manuel, Veron, Mathilde, Ghannay, Sahar, Bernard, Guillaume, Bredin, Hervé, Galibert, Olivier, Rosset, Sophie

Publikováno v: Proceedings of the First Workshop on Performance and Interpretability Evaluations of Multimodal, Multipurpose, Massive-Scale Models
First Workshop on Performance and Interpretability Evaluations of Multimodal, Multipurpose, Massive-Scale Models
First Workshop on Performance and Interpretability Evaluations of Multimodal, Multipurpose, Massive-Scale Models, Oct 2022, virtual, South Korea

International audience; Knowledge transfer between neural language models is a widely used technique that has proven to improve performance in a multitude of natural language tasks, in particular with the recent rise of large pre-trained language mod

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=od_______165::ee14f60f881e122a4ce61a9ce31635c5
https://hal.science/hal-03824597